A power distribution network operation mode adjustment method based on dijkstra algorithm

Abstract

This application provides a method for adjusting the operation mode of a distribution network based on Dijkstra's algorithm, belonging to the technical field of distribution network operation mode adjustment. The method assigns a first maintenance weight to closed switches in each branch and a second operational weight to open switches in each branch, determining the equivalent operational weight of each branch in the distribution network topology. Based on the equivalent operational weight, Dijkstra's algorithm is invoked to search for initial transfer paths from power source nodes to de-energized load nodes. A fixed offset disturbance is applied to the initial transfer paths, and Dijkstra's algorithm is iteratively invoked to generate multiple alternative transfer paths. After adjusting the switch states according to each alternative transfer path, the distribution network operation state is verified to obtain the target reconfiguration strategy for the distribution network operation mode. This application can dynamically adjust the search mechanism for transfer paths to improve the efficiency of switch operations during the distribution network operation mode adjustment process.

Description

Technical Field

[0001] This application relates to the field of power distribution network operation mode adjustment technology, and more specifically, to a power distribution network operation mode adjustment method based on Dijkstra's algorithm. Background Technology

[0002] Distribution network operation mode adjustment refers to the technical means of restoring power supply and optimizing network operation performance by reconstructing the distribution network topology by changing the opening and closing states of sectionalizing switches and tie switches in the network under operating conditions such as distribution network failure, equipment maintenance, or significant load changes. This involves transferring affected loads to other normal power supply feeders. This technology is widely used in scenarios such as fault self-healing, planned maintenance power transfer, loss reduction and energy saving, and load balancing in urban distribution networks. It is a core supporting technology for improving the reliability and economic operation of distribution network power supply. The common implementation framework in this field usually includes topology modeling, power transfer path search, safety constraint verification, and control strategy generation. Its technical value lies in achieving rapid recovery of power-loss loads with minimal switching operation costs, while ensuring that the reconstructed distribution network meets safety operation constraints such as voltage quality and equipment thermal stability.

[0003] However, current adjustments to distribution network operation modes typically rely on static, fixed-weight network topology models. The path optimization process involves only a single search to find the unique shortest path, neglecting the varying operational costs due to the current operating status of switching equipment. This often results in numerous unnecessary switching actions during the actual execution of the searched transfer path, leading to excessive switching operations, increased on-site operational risks, and low reconfiguration efficiency. Consequently, in time-sensitive scenarios such as fault repair or planned maintenance, this prolongs load outage time and hinders further improvements in distribution network reliability. Therefore, dynamically adjusting the transfer path search mechanism to improve the efficiency of switching operations during distribution network operation mode adjustments is a key challenge facing the industry. Summary of the Invention

[0004] This application provides a method for adjusting the operation mode of a distribution network based on Dijkstra's algorithm, which can dynamically adjust the search mechanism for power transfer paths to improve the efficiency of switching operations during the adjustment of the distribution network operation mode.

[0005] Firstly, this application provides a method for adjusting the operation mode of a distribution network based on Dijkstra's algorithm, wherein the method: Construct a distribution network topology model, extract node and branch relationships from the distribution network topology model to generate a distribution network topology diagram, and determine the energy loss weight of each branch in the distribution network topology diagram; A first maintenance weight is assigned to the closed switches in each branch, and a second operation weight is assigned to the open switches in each branch to form a switch state cost weight. The energy loss weight and the switch state cost weight are then fused to obtain the equivalent operating weight of each branch in the distribution network topology. Based on the equivalent weight of the operation, Dijkstra's algorithm is called to search for the initial transfer path from the power node to the power failure load node. A fixed offset perturbation is applied to the initial transfer path. Dijkstra's algorithm is iteratively called to generate multiple alternative transfer paths until the current searched transfer path is repeated with the generated alternative transfer path. After adjusting the switch status according to each alternative power transfer path, the operation status of the distribution network is checked to obtain the target reconfiguration strategy of the distribution network operation mode.

[0006] In this embodiment, a first maintenance weight is assigned to the closed switches in each branch, and a second operation weight is assigned to the open switches in each branch. The switch state cost weight specifically includes: Traverse the switching equipment of each branch in the distribution network topology model, extract the current operating status data, and classify the current operating status data into closed status and open status; A first maintenance weight is assigned to a switchgear that is in the closed state, and a second operation weight is assigned to a switchgear that is in the open state, wherein the value of the first maintenance weight is less than the value of the second operation weight. The first maintenance weight and the second operation weight are mapped to the branch where the corresponding switching device is located to obtain the switch state cost weight.

[0007] In this embodiment, the energy loss weight and the switch state cost weight are fused to obtain the equivalent operating weight of each branch in the distribution network topology diagram, specifically including: The first fusion parameter and the switching scaling factor are determined based on the preset energy loss allocation coefficient and the switching cost allocation coefficient, respectively. The loss weighting value is determined based on the energy loss weight and the first fusion parameter, and the switch weighting value is determined based on the switch state cost weight and the switch scaling factor. The loss weighting value is coupled with the switch weighting value to generate the equivalent operating weight of each branch in the distribution network topology diagram.

[0008] In this embodiment, determining the loss weighting value based on the energy loss weight and the first fusion parameter, and determining the switch weighting value based on the switch state cost weight and the switch scaling factor specifically includes: The energy loss weights are scaled according to the first fusion parameters to obtain a loss weighted value; The intermediate switch quantity is discretely quantized based on the quantization threshold coefficient set by the switch scaling factor to obtain the switch weight value.

[0009] In this embodiment, the process of searching for the initial power transfer path from the power node to the power-depleted load node using the Dijkstra algorithm based on the equivalent weights includes: Starting from the power supply node, traverse the adjacent nodes of the distribution network topology and construct the cumulative path cost of each adjacent node based on the operational equivalent weight. The cumulative path cost of all unvisited nodes is compared one by one, and the node corresponding to the boundary of the cumulative path cost is selected as the current extended node and the access status identifier is updated. Based on the access status identifier, iterative adjacency traversal is performed until the power-depleted load node is marked as accessed, at which point the search terminates and the initial transfer path is output.

[0010] In this embodiment, constructing the cumulative path cost of each adjacent node based on the equivalent running weight specifically includes: Based on the power supply branch identifier between the current extended node and each adjacent node, the weight component of the corresponding branch is extracted from the running equivalent weight and determined as the branch path cost of each adjacent node. Read the historical cumulative path cost value recorded by the current extended node, and perform step-by-step superposition of the historical cumulative path cost value and the branch path cost value to generate the cumulative path cost value of each adjacent node.

[0011] In this embodiment, applying a fixed offset perturbation to the initial transfer path and iteratively calling the Dijkstra algorithm to generate multiple alternative transfer paths, until the process terminates when the currently searched transfer path overlaps with an already generated alternative transfer path, specifically includes: Extract each power supply branch contained in the generated power supply path, and superimpose a fixed offset penalty value on the corresponding operating equivalent weight of each power supply branch to generate a perturbation topology weight map. Based on the perturbation topology weight graph, Dijkstra's algorithm is invoked to perform path search, and the currently searched transfer path is written into the candidate path set; The node sequence of the currently searched transfer path is compared one by one with the node sequences stored in the candidate path set. If there is a completely consistent matching result, the iterative search is terminated. If there is no completely consistent matching result, the currently searched transfer path is written into the candidate path set, and the process of weight perturbation and path search is returned to continue generating new paths.

[0012] In this embodiment, the process of generating a perturbation topology weight map by superimposing a fixed offset penalty value on the operating equivalent weights corresponding to each of the aforementioned electron supply branches specifically includes: Traverse each power supply branch included in the generated power supply path, and add a fixed offset penalty value to the operating equivalent weight corresponding to each power supply branch one by one to generate the disturbance weight value corresponding to each power supply branch. Each disturbance weight value is written back to the corresponding branch weight field in the topology data structure storing the distribution network topology map, keeping the equivalent operating weight of the remaining branches in the distribution network topology map unchanged, and generating a disturbance topology weight map.

[0013] In this embodiment, after adjusting the switch status according to each alternative power transfer path, the distribution network operation status is checked, and the target reconfiguration strategy for the distribution network operation mode is obtained, specifically including: The opening and closing status of sectionalizing switches and tie switches in the distribution network topology is updated based on the switching action sequence included in each alternative power transfer path. Power flow calculation is performed to extract the voltage amplitude and branch load rate of each node, and compared with the voltage safety threshold and thermal stability limit threshold to generate a verification identifier. Select qualified candidate transfer paths based on the verification identifier, construct a candidate strategy set, extract the cumulative path cost corresponding to each candidate strategy and sort them in ascending order, and select the candidate transfer path corresponding to the first position of the sequence as the target reconstruction strategy.

[0014] In this embodiment, the process of selecting candidate supply paths whose verification identifiers are qualified and constructing a candidate strategy set specifically includes: Read the voltage over-limit flag and branch overload flag in the verification identifier corresponding to each alternative power transfer path. When the voltage over-limit flag and branch overload flag of a certain alternative power transfer path are both in a negative state, the verification identifier of the alternative power transfer path is determined to be qualified. All candidate supply paths that have been verified as qualified and their corresponding switching action sequences are summarized to form a candidate strategy set.

[0015] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects: A distribution network topology model is constructed, and node and branch relationships are extracted from the model to generate a distribution network topology map. The energy loss weight of each branch in the topology map is determined. A first maintenance weight is assigned to closed switches in each branch, and a second operation weight is assigned to open switches, forming switch state cost weights. The energy loss weights and switch state cost weights are then fused to obtain the operational equivalent weights of each branch in the topology map. Based on these operational equivalent weights, the Dijkstra algorithm is invoked to search for initial transfer paths from power source nodes to de-energized load nodes. A fixed offset disturbance is applied to these initial transfer paths, and the Dijkstra algorithm is iteratively invoked to generate multiple alternative transfer paths until the currently searched transfer path overlaps with an already generated alternative transfer path. After adjusting the switch states according to each alternative transfer path, the distribution network operation status is checked to obtain the target reconfiguration strategy for the distribution network operation mode.

[0016] Therefore, this application demonstrates that the distribution network operation status can be verified by adjusting the switch status based on each alternative power transfer path, thus obtaining the target reconfiguration strategy for the distribution network operation mode. Specifically, determining the equivalent weight of the current open / closed state of the integrated switchgear provides a precise topology optimization basis that accurately represents the actual operational cost of each branch. This effectively compensates for the inherent defect of traditional static weight models, which neglect real-time equipment conditions and result in a large number of redundant switch actions in the optimization results. It guides path searching from the source to prioritize the reuse of closed switch branches and avoid unnecessary tripping operations, significantly reducing the total number of on-site switch operations and the risk of human intervention. Ultimately, while ensuring rapid and reliable restoration of power outage loads, it greatly improves the overall execution efficiency and equipment lifespan of the distribution network operation mode adjustment. By generating multiple alternative transfer paths after weight perturbation, a set of candidate strategies covering different topology reconfiguration schemes can be obtained. This effectively overcomes the inherent limitations of traditional single-step optimization mechanisms, which are prone to failure due to the single output path and easy violation of line thermal stability or node voltage safety constraints. It provides sufficient decision space for power flow verification, ensuring that the final selected target reconfiguration strategy is not only the most streamlined at the switch operation level, but also fully meets the voltage quality and equipment load rate requirements of the distribution network, effectively enhancing the engineering feasibility of the operation mode adjustment results.

[0017] In summary, the technical solution adopted in this application can dynamically adjust the search mechanism for power transfer paths to improve the efficiency of switch operations during the adjustment of distribution network operation mode. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this embodiment of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is an exemplary flowchart of a distribution network operation mode adjustment method based on Dijkstra's algorithm provided in this application; Figure 2 This is a flowchart illustrating the process of generating alternative supply paths based on the information provided in this application. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0021] This application provides a method for adjusting the operation mode of a distribution network based on Dijkstra's algorithm. The core of this method is to construct a distribution network topology model, extract node and branch relationships from the topology model to generate a distribution network topology map, and determine the energy loss weight of each branch in the topology map. A first maintenance weight is assigned to closed switches in each branch, and a second operation weight is assigned to open switches in each branch, forming switch state cost weights. The energy loss weights and switch state cost weights are then fused to obtain the equivalent operational weights of each branch in the topology map. Based on these equivalent operational weights, Dijkstra's algorithm is invoked to search for initial transfer paths from power supply nodes to de-energized load nodes. A fixed offset disturbance is applied to these initial transfer paths, and Dijkstra's algorithm is iteratively invoked to generate multiple alternative transfer paths until the currently searched transfer path overlaps with an already generated alternative transfer path. After adjusting the switch states according to each alternative transfer path, the distribution network operation state is checked to obtain the target reconfiguration strategy for the distribution network operation mode.

[0022] To better understand the above technical solutions, a detailed description of the technical solutions will be provided below in conjunction with the accompanying drawings and specific embodiments. (Refer to...) Figure 1 As shown in the figure, this is an exemplary flowchart of a distribution network operation mode adjustment method based on Dijkstra's algorithm according to an embodiment of this application. The method includes the following steps: In step S1, a distribution network topology model is constructed, and the relationships between nodes and branches are extracted from the distribution network topology model to generate a distribution network topology diagram. The energy loss weight of each branch in the distribution network topology diagram is then determined.

[0023] In practical implementation, constructing a distribution network topology model and extracting node and branch relationships from the model to generate a distribution network topology map can be achieved in the following way: First, collect the raw operating data output by the distribution network geographic information system and the distribution automation data acquisition system. This raw operating data fully includes the spatial coordinates of busbar nodes, the model and length parameters of each distribution line, the rated capacity and impedance parameters of transformers, load connection points, the installation locations of all sectionalizing switches and tie switches, and the real-time remote signaling opening and closing status. Then, perform topology element abstraction processing: abstract the substation busbars, distribution line sectionalizing points, feeder tie nodes, transformer high and low voltage connection terminals, and various load connection points into a unified set of topology nodes. Abstract the overhead / cable lines, transformer windings, and switch connecting electrical channels between adjacent nodes into a topology node set. The distribution network topology model is constructed by first creating a set of branch paths and then recording the node numbers at both ends, line resistance and reactance, bound switch numbers, and real-time operating status of each branch. A node-branch correlation matrix is ​​generated based on this set of nodes and branches. A distribution network topology model is then constructed using the node set, branch set, node-branch correlation matrix, all electrical equipment parameters, and real-time operating status. This model serves as a complete underlying data carrier integrating all static parameters and real-time operating conditions of the distribution network, storing complete information on all nodes, branches, switches, loads, and transformers. After completing the distribution network topology model, the connectivity relationships between nodes and branches are extracted. Redundant electrical parameters such as equipment capacity and impedance are removed, retaining only node numbers and branch connection correspondences. This generates a distribution network topology graph specifically for the improved Dijkstra's Shortest Path First Algorithm (DSPF algorithm) to optimize the transfer path. This distribution network topology graph is a lightweight weighted graph theory carrier, retaining only the connectivity logic required for path calculation. All subsequent branch weight assignments and multi-path iterative search operations are completed based on this topology graph.

[0024] It should be noted that, in this application, the distribution network topology model refers to a global power grid data model that integrates all electrical parameters and real-time operating status of buses, lines, transformers, switches, and loads; a node refers to the electrical connection point of two or more branches in the distribution network; a branch refers to the electrical path connecting two adjacent nodes in the distribution network topology diagram; the distribution network topology diagram refers to a weighted graph theory model formed by extracting node and branch connection relationships from the distribution network topology model and assigning branch weights; and the topology data structure refers to the carrier in a computer program that stores the node number, branch identifier, weight value, and predecessor node of the distribution network topology diagram. In specific implementation, the energy loss weight of each branch in the distribution network topology can be determined in the following way: First, obtain the line parameters of each branch in the distribution network topology, including at least the branch resistance and branch length; second, based on the load data of each node in the topology and the line parameters, use the forward-backward power flow calculation method to solve for the current amplitude of each branch during the preset operating period; then, according to the current amplitude and branch resistance of each branch, calculate the current loss weight of each branch according to the formula... Calculate the energy loss of each branch during the operating period, where, Let i be the energy loss of the i-th branch. Let be the current amplitude of the i-th branch. Let be the resistance value of the i-th branch, and T be the operating time. Finally, the ratio of the energy loss of each branch to the total energy loss of all branches is determined as the energy loss weight of that branch. ,in, Let i be the energy loss of the i-th branch. Let N be the energy loss of the j-th branch; N is the total number of branches in the distribution network topology diagram, and j is the traversal index number of the branch.

[0025] It should be noted that, in this application, the energy loss weight refers to a quantitative indicator of the degree of active power loss generated by the distribution network branch during the power transmission process.

[0026] In step S2, a first maintenance weight is assigned to the closed switches in each branch, and a second operation weight is assigned to the open switches in each branch to form a switch state cost weight. The energy loss weight and the switch state cost weight are then fused to obtain the equivalent operating weight of each branch in the distribution network topology diagram.

[0027] In this embodiment, a first maintenance weight is assigned to the closed switches in each branch, and a second operation weight is assigned to the open switches in each branch. The switch state cost weights can be formed in the following way: Traverse the switching equipment of each branch in the distribution network topology model, extract the current operating status data, and classify the current operating status data into closed status and open status; A first maintenance weight is assigned to a switchgear that is in the closed state, and a second operation weight is assigned to a switchgear that is in the open state, wherein the value of the first maintenance weight is less than the value of the second operation weight. The first maintenance weight and the second operation weight are mapped to the branch where the corresponding switching device is located to obtain the switch state cost weight.

[0028] In practical implementation, firstly, a graph traversal algorithm is used to traverse the nodes and connecting edges in the distribution network topology model, identifying the switching equipment contained in each branch, and reading the remote signaling signals collected in real time by the distribution automation system to extract the current operating status data. Based on a preset state determination logic, the current operating status data is binarized and classified, classifying signals indicating circuit conduction as closed and signals indicating circuit disconnection as open. Then, a differentiated weight assignment operation is performed on the classified switching equipment, assigning a first maintenance weight to switching equipment in the closed state and a second operation weight to switching equipment in the open state. Based on the distribution network reconfiguration operation cost model, the value range of the first maintenance weight is set to 0.1 to 0.3, and the value range of the second operation weight is set to 0.7 to 0.9, ensuring that the value of the first maintenance weight is strictly less than the value of the second operation weight. Finally, the assigned first maintenance weight and second operation weight are mapped to the corresponding branch where the switching equipment is located, and weight aggregation calculation is performed using the branch as the basic unit to obtain the switch state cost weight. The specific execution process is as follows: extract the weight values ​​corresponding to all switching devices in the target branch, use the cumulative calculation model to sum the weight values ​​of each item in the same branch, use the summation result as the final switching state cost weight of the branch, and output the topology weight matrix containing the cost weight of each branch.

[0029] It should be noted that, in this application, a closed switch refers to a switchgear in the distribution network that is in a circuit-conducting state and is used to maintain the normal power transmission function of the current branch; the first maintenance weight is a relatively small weight parameter that quantifies the operational cost required to maintain the current conducting state of the switchgear; a closed switch refers to a switchgear in the distribution network that is in a circuit-disconnecting state and is used to block the power transmission function of the current branch; the second operation weight is a relatively large weight parameter that quantifies the operational cost required to change the current disconnected state of the switchgear; a closed state refers to the operating state in which the internal contacts of the switchgear are closed to achieve physical conduction of the electrical circuit of the branch; a closed state refers to the operating state in which the internal contacts of the switchgear are separated to achieve physical disconnection of the electrical circuit of the branch; and the switch state cost weight refers to the branch-level weight parameter of the total operational cost of adjusting the state of all switchgear in a specific branch of the distribution network.

[0030] In this embodiment, the energy loss weight and the switch state cost weight are fused to obtain the equivalent operating weight of each branch in the distribution network topology diagram. This can be achieved by the following steps: The first fusion parameter and the switching scaling factor are determined based on the preset energy loss allocation coefficient and the switching cost allocation coefficient, respectively. The loss weighting value is determined based on the energy loss weight and the first fusion parameter, and the switch weighting value is determined based on the switch state cost weight and the switch scaling factor. The loss weighting value is coupled with the switch weighting value to generate the equivalent operating weight of each branch in the distribution network topology diagram.

[0031] In practical implementation, firstly, the energy loss allocation coefficient is used to characterize the relative importance of energy loss factors in the branch operation evaluation system, and the switch cost allocation coefficient is used to characterize the relative importance of switch state switching cost factors. Specifically, the values ​​of the energy loss allocation coefficient and the switch cost allocation coefficient are pre-set according to the operational requirements of the distribution network, and their sum is normalized to one, meaning that the energy loss allocation coefficient and the switch cost allocation coefficient satisfy a complementary constraint relationship. In the normal operation optimization scenario, the energy loss allocation coefficient is set to be greater than the switch cost allocation coefficient to highlight the optimization goal of reducing line losses; in the fault recovery or maintenance operation scenario, the switch cost allocation coefficient is set to be greater than the energy loss allocation coefficient to suppress the equipment risks caused by frequent switching operations. Subsequently, the energy loss allocation coefficient is directly mapped to the first fusion parameter, and the switch cost allocation coefficient is directly mapped to the switch scaling coefficient; this switch scaling coefficient is only used for linear product scaling with the switch state cost weight, and a separate quantization threshold coefficient is set. This quantization threshold coefficient is specifically used to construct the boundary threshold of the discrete quantization interval, and is independent of the switch scaling coefficient, with no functional reuse. Then, after obtaining the first fusion parameter and the quantization threshold coefficient, weighted operations are performed on the energy loss weight and the switch state cost weight, respectively. Specifically, the energy loss weight corresponding to each branch in the distribution network topology diagram is multiplied by the first fusion parameter, and the product result is the loss weight value corresponding to that branch. The energy loss weight reflects the degree of active power loss of the branch under the current power flow distribution, and its value is determined by the product of the branch resistance parameter and the square of the current amplitude flowing through the branch. The first fusion parameter acts as a scaling factor to apply scenario adaptation adjustment to the energy loss weight. Simultaneously, the switch state cost weight corresponding to each branch in the distribution network topology diagram is multiplied by the quantization threshold coefficient, and the product result is the switch weight value corresponding to that branch. The switch state cost weight reflects the operational cost required for the switching equipment on the branch to switch from the current state to the target state, including the limit on the number of switching actions, the mechanical life loss of the equipment, and the operation timing constraints. The quantization threshold coefficient acts as a scaling factor to apply scenario adaptation adjustment to the switch state cost weight. Finally, the coupling operation employs a branch-by-branch linear superposition method, which involves summing the loss weight value corresponding to each branch with the switching weight value corresponding to that branch. The summation result is the operational equivalent weight of that branch. The operational equivalent weight comprehensively reflects the overall operational quality of that branch in the current operating scenario, balancing energy loss level and switching operation cost.A larger equivalent weight value indicates a higher overall operating cost for that branch. During the Dijkstra's algorithm's minimum-cost path optimization, the algorithm will try to avoid this branch and will not prioritize its construction as a transfer channel. Conversely, a smaller equivalent weight value indicates a lower overall operating cost for that branch. The algorithm is more inclined to select this branch for power supply path construction, adapting to existing switch and load conditions and prioritizing its current operating status. After traversing all branches in the distribution network topology and performing the above coupling operations for each branch, the complete equivalent weights for each branch are output.

[0032] It should be noted that, in this application, fusion refers to the process of combining two or more independent evaluation indicators into a unified comprehensive indicator according to preset rules; the energy loss allocation coefficient refers to a preset proportional parameter representing the relative importance of energy loss factors in the comprehensive evaluation system of branch operation; the switch cost allocation coefficient refers to a preset proportional parameter representing the relative importance of switch state switching cost factors in the comprehensive evaluation system of branch operation; the first fusion parameter refers to the computational parameter obtained by mapping the energy loss allocation coefficient and used to apply proportional scaling adjustment to the energy loss weight; the loss weighting value is a weighted result reflecting the relative degree of active power loss in the branch; the switch weighting value is a weighted result reflecting the relative degree of switch operation cost in the branch; the operational equivalent weight refers to an evaluation indicator of the overall operational cost quality of the branch; the switch scaling coefficient refers to a linear scaling parameter generated by mapping the switch cost allocation coefficient; and the quantization threshold coefficient refers to an independently preset dedicated parameter, used only as the interval boundary threshold for discrete classification of switch quantities.

[0033] In addition, in this embodiment, determining the loss weighting value based on the energy loss weight and the first fusion parameter, and determining the switch weighting value based on the switch state cost weight and the switch scaling factor, can be achieved by the following steps: The energy loss weights are scaled according to the first fusion parameters to obtain a loss weighted value; The intermediate switch quantity is discretely quantized based on the quantization threshold coefficient set by the switch scaling factor to obtain the switch weight value.

[0034] In specific implementation, firstly, the energy loss weight and the first fusion parameter are used as input data, and a linear scaling model is used to perform scaling operations. The first fusion parameter, as a scaling factor, is set to a closed interval between zero and one, theoretically to prevent excessive weight amplification that could lead to an imbalance in the optimization objective. The processing logic involves directly multiplying the energy loss weight and the first fusion parameter. When the first fusion parameter approaches 1, the original weight characteristics are retained; when it approaches zero, the impact of loss is suppressed, and the final output is a scaled loss-weighted value. Then, the switch state cost weight is multiplied by the switch scaling factor to obtain the intermediate switch quantity. Then, using an independently set quantization threshold coefficient as the hierarchical boundary, discrete state quantization is performed on the intermediate switch quantity. The quantization threshold coefficient is a dedicated interval division parameter that does not participate in the linear scaling operation, and its value is determined based on the number of physical action positions of the distribution network switchgear. The processing logic involves comparing the continuous switch state cost weights with the threshold values ​​set by the quantization threshold coefficient for each level, mapping the cost weights falling within a specific threshold interval to the corresponding integer discrete state codes, and finally outputting the switch weighted value representing the switch action level.

[0035] It should be noted that, in this application, scaling refers to the process of proportionally scaling the continuous weight values; discrete state quantization refers to the conversion mechanism of mapping continuous values ​​to integer state codes to match the discrete action level characteristics of physical equipment.

[0036] In step S3, based on the running equivalent weight, the Dijkstra algorithm is called to search for an initial transfer path from the power node to the power failure load node. A fixed offset perturbation is applied to the initial transfer path, and the Dijkstra algorithm is iteratively called to generate multiple alternative transfer paths until the currently searched transfer path is the same as the generated alternative transfer path, and then the process terminates.

[0037] In this embodiment, the initial power transfer path from the power node to the power-depleted load node can be searched using the Dijkstra algorithm based on the equivalent weights obtained from the operation, through the following steps: Starting from the power supply node, traverse the adjacent nodes of the distribution network topology and construct the cumulative path cost of each adjacent node based on the operational equivalent weight. The cumulative path cost of all unvisited nodes is compared one by one, and the node corresponding to the boundary of the cumulative path cost is selected as the current extended node and the access status identifier is updated. Based on the access status identifier, iterative adjacency traversal is performed until the power-depleted load node is marked as accessed, at which point the search terminates and the initial transfer path is output.

[0038] In practice, firstly, the power supply node with power supply capability is set as the initial source node of the Dijkstra algorithm, and its initial cumulative path cost is set to zero. Starting from this initial source node, the directly connected adjacent nodes in the distribution network topology are extracted, and the equivalent operating weights of the lines connecting them are read. The cumulative path cost of the power supply node is added to the equivalent operating weights of the corresponding lines to obtain the cumulative path cost of each adjacent node. For nodes that are not directly connected, their cumulative path cost is initialized to infinity. Then, among all nodes in the distribution network topology, the set of nodes with the access status marked as unvisited is extracted. The cumulative path cost of all nodes in this set is traversed, and the boundary of the minimum cumulative path cost is selected by numerical comparison. The node corresponding to this lower limit is determined as the current expansion node, and its access status is updated to visited to prevent duplicate calculations in subsequent iterations. At the same time, the cumulative path cost of its unvisited adjacent nodes is updated based on the current expansion node; if the newly calculated path cost is less than the original value, it is replaced. Finally, using the current extended node as the new starting point, continue to extract its unvisited adjacent nodes and calculate the cumulative path cost. Repeat the node comparison, extended node filtering, and access status flag update operations to form a closed-loop iterative mechanism. After each iteration, check the access status flag of the power-depleted load node; if a power-depleted load node is detected as marked as visited, immediately terminate the iterative search. Backtrack from the power-depleted load node to its predecessor node to construct a complete node sequence from the power source node to the power-depleted load node; the output is the initial power transfer path.

[0039] It should be noted that in this application, Dijkstra's algorithm refers to the Dijkstra algorithm, used to calculate the shortest path for a single source in a weighted graph to search for the optimal transfer channel from the power source to the de-energized load; a power source node refers to a distribution network node with continuous power supply capability; a de-energized load node refers to a terminal electrical equipment or distribution transformer node whose power supply is interrupted due to a distribution network fault and needs to be restored through a transfer path; an adjacent node refers to a neighboring grid node in the distribution network topology graph that is directly connected to the currently examined node through a physical line and has a power transmission connectivity relationship; the cumulative path cost refers to the sum of the equivalent weights of all lines on the path from the power source node to the current node; all unvisited nodes refer to nodes that have not yet been selected as the current expansion node during the path search process; the cumulative path cost boundary refers to the boundary of all unvisited nodes. Among the cumulative path cost values ​​of nodes, the minimum cost value is selected through numerical comparison; the current expansion node refers to the central grid node with the minimum cumulative path cost value in a single iteration, which is selected for outward expansion search and updating the cost values ​​of adjacent nodes; the access status identifier refers to the logical marker recording whether each node in the distribution network topology has been determined as the current expansion node; the adjacency traversal refers to the search operation that, with the current expansion node as the center, filters and calculates the cost values ​​of its directly connected nodes one by one along the physical connection relationship of the distribution network topology; the visited state refers to the marked state after a node has been selected as the current expansion node and its adjacent node cost values ​​have been updated; the initial transfer path refers to the initial power transmission channel from the power source node to the de-energized load node with the minimum cumulative path cost value, obtained by searching using the Dijkstra algorithm.

[0040] In addition, in this embodiment, the cumulative path cost of each adjacent node can be constructed based on the equivalent running weight using the following steps: Based on the power supply branch identifier between the current extended node and each adjacent node, the weight component of the corresponding branch is extracted from the running equivalent weight and determined as the branch path cost of each adjacent node. Read the historical cumulative path cost value recorded by the current extended node, and perform step-by-step superposition of the historical cumulative path cost value and the branch path cost value to generate the cumulative path cost value of each adjacent node.

[0041] In practical implementation, firstly, in the distribution network topology diagram, all directly connected neighboring nodes are identified with the current extended node as the center. For each neighboring node, the power supply branch identifier corresponding to the physical line connecting the current extended node and that neighboring node is read. This power supply branch identifier serves as an index key, and an exact match query is performed in the pre-built operational equivalent weight database to extract the operational equivalent weight value corresponding to that specific branch; this value is the weight component. This weight component is directly assigned to the corresponding neighboring node as the single-step branch path cost required to reach that neighboring node from the current extended node. All neighboring nodes are traversed to determine the branch path cost value of each neighboring node one by one. Then, the historical cumulative path cost value calculated and saved by the current extended node in the previous iteration is read from the node status record table; this value represents the minimum cumulative cost to reach the current extended node from the power supply node. For each neighboring node, its corresponding branch path cost value is arithmetically added to the historical cumulative path cost value of the current extended node, i.e., a step-by-step superposition operation is performed. The sum of these values ​​represents the temporary cumulative path cost from the power node to the adjacent node via the current extended node. This temporary cumulative path cost is compared with the existing cumulative path cost of the adjacent node, and if the temporary value is smaller, it is updated and replaced, ultimately generating the latest cumulative path cost for each adjacent node.

[0042] It should be noted that, in this application, the power supply branch identifier refers to the physical power supply line that connects two adjacent nodes in the distribution network topology; the weight component refers to a single numerical parameter representing the magnitude of the comprehensive operating cost of a specific power supply branch; the branch path cost refers to the single-step transmission cost required to reach a certain adjacent node from the current extended node; the historical cumulative path cost refers to the minimum cumulative transmission cost recorded from the power supply node to the current extended node in the previous iteration; and the step-by-step superposition refers to the computational mechanism of arithmetically adding the historical cumulative path cost of the current node to the single-step branch path cost to calculate the cumulative cost of the next node.

[0043] Preferably, in this embodiment, a fixed offset perturbation is applied to the initial transfer path, and the Dijkstra algorithm is iteratively called to generate multiple alternative transfer paths until the currently searched transfer path overlaps with an already generated alternative transfer path, at which point the process terminates. Figure 2 As shown in the figure, this is a schematic flowchart of generating alternative supply paths in some embodiments of this application. In this embodiment, generating alternative supply paths can be achieved by the following steps: In step S31, each power supply branch included in the generated power supply path is extracted, and a fixed offset penalty value is superimposed on the operating equivalent weight corresponding to each power supply branch to generate a disturbance topology weight map. In step S32, the Dijkstra algorithm is invoked to perform path search based on the perturbation topology weight graph, and the currently searched transfer path is written into the candidate path set. In step S33, the node sequence of the currently searched transfer path is compared one by one with each node sequence stored in the candidate path set. If there is a completely consistent matching result, the iterative search is terminated. If there is no completely consistent matching result, the currently searched transfer path is written into the candidate path set, and the process of weight perturbation and path search is returned to continue generating new paths.

[0044] In practice, firstly, all power supply branches in the currently generated transfer paths are extracted, and their corresponding operational equivalent weights are obtained. The operational equivalent weights reflect the comprehensive operational cost of the power supply branch in the distribution network, including line impedance, load factor, and loss indicators. A fixed offset penalty value is superimposed on the operational equivalent weight of each extracted power supply branch. This fixed offset penalty value is a preset positive real constant, ranging from 10% to 30% of the original operational equivalent weight, ensuring that the penalty effectively guides subsequent path searches away from the selected branches without excessively distorting the true weight distribution of the network topology. The weights of each branch after the penalty is added, along with the weights of the remaining branches that are not penalized, are written into the topology weight matrix to generate a perturbed topology weight map. Then, based on the perturbed topology weight map, the Dijkstra algorithm is invoked to perform the shortest path search, with the transfer origin node in the distribution network as the search source node and the target receiving node as the search destination node. The specific execution flow of Dijkstra's algorithm in this scheme is as follows: First, initialize the distance label of the search source node to zero, and the distance labels of all other nodes to positive infinity, marking all nodes as unvisited. Then, select the node with the smallest distance label from the unvisited nodes as the current node, traverse all its neighboring nodes, and calculate the cumulative path cost from the current node to each neighboring node based on the weight value of the corresponding branch in the perturbed topology weight graph. If the cumulative path cost is less than the current distance label of the neighboring node, update the label and record the predecessor node. Repeat the above selection and relaxation operations until the target receiving node is marked as visited. After the search is completed, backtrack from the target receiving node along the predecessor node chain to the search source node to obtain the current transfer path, and write the transfer path into the candidate path set. Finally, obtain all the nodes traversed by the current transfer path, and arrange them in order from the transfer starting node to the target receiving node to form the current node sequence. The current node sequence is compared one by one with the node sequences corresponding to each of the stored transfer paths in the candidate path set. The comparison rules are as follows: check whether the lengths of the two node sequences are equal, and if the lengths are equal, compare whether the node numbers at corresponding positions are completely identical. If there is a stored node sequence that is completely identical to the current node sequence in both length and node numbers, it is determined that the currently searched transfer path is a duplicate of an existing candidate path, the iterative search is terminated immediately, and all currently stored transfer paths in the candidate path set are output as the final result. If no completely identical match is found, the current transfer path is retained in the candidate path set, and the next round of perturbation and search iterations is performed.

[0045] It should be noted that, in this application, fixed offset perturbation refers to the topology weight adjustment operation that applies a fixed incremental penalty to the operating equivalent weight of the used branch to guide the path search direction away from the selected branch; a transfer path refers to a complete power transmission path in the distribution network consisting of several power supply branches connected sequentially from the transfer start node to the target power receiving node; a power supply branch refers to the smallest line unit in the distribution network topology that connects two adjacent nodes and undertakes the function of power transmission; a fixed offset penalty value refers to a positive real constant that is uniformly superimposed on the operating equivalent weight of the used power supply branches during the perturbation process; a perturbation topology weight graph refers to the distribution network topology data structure containing the weights of all nodes and branches after applying the fixed offset penalty value to the used power supply branches; path search refers to the process of solving for the path with the minimum cumulative cost from the start node to the end node in the topology weight graph based on Dijkstra's algorithm; a candidate path set refers to an ordered data set that stores each unique transfer path generated during the iterative search process; and a node sequence refers to an ordered list consisting of all node numbers arranged sequentially from the start node to the end node in a transfer path.

[0046] In addition, in this embodiment, the generation of the perturbation topology weight map by superimposing a fixed offset penalty value on the operating equivalent weights corresponding to each of the electron supply branches can be achieved by the following steps: Traverse each power supply branch included in the generated power supply path, and add a fixed offset penalty value to the operating equivalent weight corresponding to each power supply branch one by one to generate the disturbance weight value corresponding to each power supply branch. Each disturbance weight value is written back to the corresponding branch weight field in the topology data structure storing the distribution network topology map, while keeping the equivalent operating weights of the remaining branches in the distribution network topology map unchanged, thus generating a disturbance topology weight map.

[0047] In practice, firstly, assuming at least one transfer path has been generated in the distribution network topology diagram, the system sequentially extracts all power supply branches included in the path along its topological connection sequence. For each power supply branch, the system reads its calculated equivalent operating weight under the current operating state. This equivalent operating weight reflects the branch's comprehensive indicators of load capacity, line loss level, and voltage deviation under normal power supply conditions. Then, the system superimposes a preset fixed offset penalty value onto this equivalent operating weight. The fixed offset penalty value is a real number greater than zero, and its value is set based on the branch failure rate statistics in the historical operating data of the distribution network, ranging from 10% to 30% of the original equivalent operating weight. The superposition operation uses a linear addition model, directly adding the equivalent operating weight to the fixed offset penalty value. The sum is used as the disturbance weight value corresponding to that power supply branch. Finally, after calculating the disturbance weight values ​​for all power supply branches, each disturbance weight value is written back to the weight field of the corresponding branch in the distribution network topology diagram. Specifically, based on the unique branch identifier of each power supply branch in the distribution network topology diagram, the system locates the field where the weight of that branch is stored in the topology data structure. It then replaces the existing stored operational equivalent weight with the newly generated disturbance weight value. Simultaneously, the operational equivalent weights of the remaining branches in the distribution network topology diagram not included in the current power transfer path remain unchanged, and the system does not modify them. After all write-back operations are completed, the system outputs an updated distribution network topology diagram. In this updated topology diagram, the weights of some branches have been replaced by disturbance weight values, while the weights of the remaining branches remain unchanged. This updated topology diagram is defined as the disturbance topology weight diagram.

[0048] It should be noted that, in this application, the disturbance weight value refers to the weight obtained by superimposing a fixed offset penalty value on the operating equivalent weight of the power supply branch; the weight field refers to the storage location in the distribution network topology data structure used to store the weight values ​​of each branch.

[0049] In step S4, after adjusting the switch status according to each alternative power transfer path, the operation status of the distribution network is checked to obtain the target reconfiguration strategy of the distribution network operation mode.

[0050] In this embodiment, after adjusting the switch status according to each alternative power transfer path and verifying the distribution network operation status, the target reconfiguration strategy for the distribution network operation mode can be achieved through the following steps: The opening and closing status of sectionalizing switches and tie switches in the distribution network topology is updated based on the switching action sequence included in each alternative power transfer path. Power flow calculation is performed to extract the voltage amplitude and branch load rate of each node, and compared with the voltage safety threshold and thermal stability limit threshold to generate a verification identifier. Select qualified candidate transfer paths based on the verification identifier, construct a candidate strategy set, extract the cumulative path cost corresponding to each candidate strategy and sort them in ascending order, and select the candidate transfer path corresponding to the first position of the sequence as the target reconstruction strategy.

[0051] In practice, firstly, the corresponding switch action sequence is extracted from each alternative power transfer path. This sequence records the numbers and target states of all switchgear required to activate the path. Based on this sequence, control commands are issued to the distribution network topology model to switch the corresponding sectionalizing switches from closed to open and the corresponding tie switches from open to closed, thus dynamically updating the topology. Based on the updated distribution network topology model, the Newton-Raphson method is used to perform power flow calculations. Input data includes the active and reactive power injection values ​​of each power source node, the active and reactive load demands of each load node, and the impedance parameters of each branch. The Jacobian matrix correction equation is solved iteratively until the power imbalance at each node converges to below a preset accuracy threshold. The voltage amplitude of each node and the load rate of each branch after convergence are then extracted. The voltage amplitude of each node is compared with a preset voltage safety threshold, and the load rate of each branch is compared with a preset thermal stability limit threshold. The voltage safety threshold is set to 95% to 105% of the nominal voltage, and the thermal stability limit threshold is set to 100% of the branch's rated current carrying capacity. If the voltage amplitude of all nodes is within the voltage safety threshold range and the load rate of all branches does not exceed the thermal stability limit threshold, a verification mark is generated for the candidate power transfer path as qualified; otherwise, a verification mark is generated as unqualified. Then, all candidate power transfer paths are traversed, and the verification mark corresponding to each path is read. Only paths with qualified verification marks are retained and added to the candidate strategy set, while all paths with unqualified verification marks are removed. The cumulative path cost value corresponding to each candidate strategy is extracted from the candidate strategy set. This cumulative path cost value is the sum of the equivalent weights of all branches from the power supply node to the power failure load node calculated during the Dijkstra algorithm search process for the candidate power transfer path. All cumulative path cost values ​​in the candidate strategy set are sorted in ascending order, that is, the cumulative path cost value with the smallest value is placed at the beginning of the sequence, and the cumulative path cost value with the largest value is placed at the end of the sequence. The candidate transfer path corresponding to the first position of the sequence after ascending order is selected as the target reconfiguration strategy. This target reconfiguration strategy is the optimal distribution network reconfiguration scheme that satisfies all safe operation constraints and minimizes the overall transmission cost.

[0052] It should be noted that, in this application, "switch state" refers to the physical operating state of the sectionalizing switches and tie switches in the distribution network, which are currently closed or open; "distribution network operation state verification" refers to the process of calculating the power flow of the distribution network after the implementation of alternative power transfer paths and verifying whether the safe operation constraints are met; "distribution network operation mode" refers to the overall operation configuration scheme of each power source, load, and switching equipment in the distribution network under a specific topology; "switch action sequence" refers to the ordered set of all switching equipment numbers and their corresponding target operation states required to achieve the conduction of alternative power transfer paths; "sectioning switches and tie switches" refer to the closed or open states of the sectionalizing switches used to isolate line sections and the tie switches used to connect different feeders in the distribution network; "node voltage magnitude" refers to the magnitude of the voltage phasor at each electrical node in the distribution network topology after the power flow calculation converges; and "branch load factor" refers to the current actual transmission power of the branch. The ratio between the rated transmission capacity of the branch and the voltage safety threshold refers to the upper and lower limits of the allowable fluctuation of voltage amplitude at each node of the distribution network; the thermal stability limit threshold refers to the maximum current carrying capacity that the branch can withstand under continuous operation conditions; the verification mark refers to the binary judgment mark generated after the power flow safety verification, which is divided into qualified mark and unqualified mark; the qualified mark means that the candidate transfer path meets all safety constraints of node voltage and branch thermal stability, and the unqualified mark means that the path has voltage over-limit or branch overload problems; the candidate strategy set refers to the set of reconfiguration schemes that are physically feasible and consist of all candidate transfer paths with qualified verification marks; ascending order refers to the sorting operation of the cumulative path cost corresponding to each candidate strategy in the candidate strategy set in ascending order of value; the target reconfiguration strategy refers to the optimal distribution network reconfiguration scheme with the minimum cumulative path cost and that meets all safety operation constraints selected from the candidate strategy set.

[0053] In addition, in this embodiment, the process of selecting candidate supply paths with qualified verification identifiers and constructing a candidate strategy set can be achieved through the following steps: Read the voltage over-limit flag and branch overload flag in the verification flag corresponding to each alternative power transfer path. When the voltage over-limit flag and branch overload flag of a certain alternative power transfer path are both in a negative state, the verification flag of the alternative power transfer path is determined to be qualified. All candidate supply paths that have been verified as qualified and their corresponding switching action sequences are summarized to form a candidate strategy set.

[0054] In specific implementation, first, traverse all alternative transfer paths, and read the voltage out-of-limit flag and branch overload flag included in the check mark corresponding to each alternative transfer path one by one. The voltage out-of-limit flag is a binary logic variable. When there is any node voltage amplitude exceeding the voltage safety threshold in the power flow calculation result, this flag is set to true; when all node voltage amplitudes fall within the voltage safety threshold interval, this flag is set to false. The branch overload flag is also a binary logic variable. When there is any branch load rate exceeding the thermal stability limit threshold, this flag is set to true; when all branch load rates do not exceed the thermal stability limit threshold, this flag is set to false. The system performs a joint judgment on each alternative transfer path: only when both the voltage out-of-limit flag and the branch overload flag are in a negative state, the check mark corresponding to the alternative transfer path is determined to be qualified; if either the voltage out-of-limit flag or the branch overload flag is true, the check mark is determined to be unqualified, and the alternative transfer path is discarded. After completing the judgment of all alternative transfer paths, all alternative transfer paths with qualified check marks are extracted. For each alternative transfer path with a qualified check mark, call its matching switch action sequence, where the switch action sequence records all switch equipment numbers and target opening / closing operation instructions required to realize the conduction of the transfer path. Bind a single path and its corresponding switch action sequence into a complete candidate strategy record, and the record content includes the path node sequence and the matching switch action sequence; aggregate all candidate strategy records that meet the conditions and store them in a candidate strategy set in the form of a list.

[0055] It should be noted that in the present application, the voltage out-of-limit flag refers to a binary logic judgment result that marks whether there is a node voltage amplitude exceeding the voltage safety threshold range after the alternative transfer path is implemented; the branch overload flag refers to a binary logic judgment result that marks whether there is a branch load rate exceeding the thermal stability limit threshold after the alternative transfer path is implemented; the voltage out-of-limit flag is a binary logic quantity used to characterize whether the node voltage exceeds the threshold in the power flow calculation result of the alternative transfer path, a true value represents that there is a node voltage out-of-limit, and a false value represents that all node voltages meet the safety interval requirements. The branch overload flag is a binary logic quantity used to characterize whether the branch load exceeds the limit after the power flow calculation of the alternative transfer path, a true value represents that the branch is overloaded, and a false value represents that all branch loads do not exceed the thermal stability limit.

[0056] Therefore, this application demonstrates that the distribution network operation status can be verified by adjusting the switch status based on each alternative power transfer path, thus obtaining the target reconfiguration strategy for the distribution network operation mode. Specifically, determining the equivalent weight of the current open / closed state of the integrated switchgear provides a precise topology optimization basis that accurately represents the actual operational cost of each branch. This effectively compensates for the inherent defect of traditional static weight models, which neglect real-time equipment conditions and result in a large number of redundant switch actions in the optimization results. It guides path searching from the source to prioritize the reuse of closed switch branches and avoid unnecessary tripping operations, significantly reducing the total number of on-site switch operations and the risk of human intervention. Ultimately, while ensuring rapid and reliable restoration of power outage loads, it greatly improves the overall execution efficiency and equipment lifespan of the distribution network operation mode adjustment. By generating multiple alternative transfer paths after weight perturbation, a set of candidate strategies covering different topology reconfiguration schemes can be obtained. This effectively overcomes the inherent limitations of traditional single-step optimization mechanisms, which are prone to failure due to the single output path and easy violation of line thermal stability or node voltage safety constraints. It provides sufficient decision space for power flow verification, ensuring that the final selected target reconfiguration strategy is not only the most streamlined at the switch operation level, but also fully meets the voltage quality and equipment load rate requirements of the distribution network, effectively enhancing the engineering feasibility of the operation mode adjustment results.

[0057] In summary, the technical solution adopted in this application can dynamically adjust the search mechanism for power transfer paths to improve the efficiency of switch operations during the adjustment of distribution network operation mode.

[0058] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0059] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-Erasable Programmable Read-Only Memory (EEPROM), compactdisc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

[0060] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

Claims

1. A method for adjusting the operation mode of a distribution network based on Dijkstra's algorithm, characterized in that, The method includes: Construct a distribution network topology model, extract node and branch relationships from the distribution network topology model to generate a distribution network topology diagram, and determine the energy loss weight of each branch in the distribution network topology diagram; A first maintenance weight is assigned to the closed switches in each branch, and a second operation weight is assigned to the open switches in each branch to form a switch state cost weight. The energy loss weight and the switch state cost weight are then fused to obtain the equivalent operating weight of each branch in the distribution network topology. Based on the equivalent weight of the operation, Dijkstra's algorithm is called to search for the initial transfer path from the power node to the power failure load node. A fixed offset perturbation is applied to the initial transfer path. Dijkstra's algorithm is iteratively called to generate multiple alternative transfer paths until the current searched transfer path is repeated with the generated alternative transfer path. After adjusting the switch status according to each alternative power transfer path, the operation status of the distribution network is checked to obtain the target reconfiguration strategy of the distribution network operation mode.

2. The method for adjusting the operation mode of a distribution network based on Dijkstra's algorithm as described in claim 1, characterized in that, A first maintenance weight is assigned to the closed switches in each branch, and a second operation weight is assigned to the open switches in each branch, forming the switch state cost weight, which specifically includes: Traverse the switching equipment of each branch in the distribution network topology model, extract the current operating status data, and classify the current operating status data into closed status and open status; A first maintenance weight is assigned to a switchgear that is in the closed state, and a second operation weight is assigned to a switchgear that is in the open state, wherein the value of the first maintenance weight is less than the value of the second operation weight. The first maintenance weight and the second operation weight are mapped to the branch where the corresponding switching device is located to obtain the switch state cost weight.

3. The method for adjusting the operation mode of a distribution network based on Dijkstra's algorithm as described in claim 1, characterized in that, The energy loss weight and the switch state cost weight are fused to obtain the equivalent operating weight of each branch in the distribution network topology diagram, specifically including: The first fusion parameter and the switching scaling factor are determined based on the preset energy loss allocation coefficient and the switching cost allocation coefficient, respectively. The loss weighting value is determined based on the energy loss weight and the first fusion parameter, and the switch weighting value is determined based on the switch state cost weight and the switch scaling factor. The loss weighting value is coupled with the switch weighting value to generate the equivalent operating weight of each branch in the distribution network topology diagram.

4. The method for adjusting the operation mode of a distribution network based on Dijkstra's algorithm as described in claim 3, characterized in that, Determining the loss weighting value based on the energy loss weight and the first fusion parameter, and determining the switch weighting value based on the switch state cost weight and the switch scaling factor, specifically includes: The energy loss weights are scaled according to the first fusion parameters to obtain a loss weighted value; The intermediate switch quantity is discretely quantized based on the quantization threshold coefficient set by the switch scaling factor to obtain the switch weight value.

5. The method for adjusting the operation mode of a distribution network based on Dijkstra's algorithm as described in claim 1, characterized in that, The search for the initial power transfer path from the power node to the power-out load node based on the aforementioned equivalent weights using the Dijkstra algorithm specifically includes: Starting from the power supply node, traverse the adjacent nodes of the distribution network topology and construct the cumulative path cost of each adjacent node based on the operational equivalent weight. The cumulative path cost of all unvisited nodes is compared one by one, and the node corresponding to the boundary of the cumulative path cost is selected as the current extended node and the access status identifier is updated. Based on the access status identifier, iterative adjacency traversal is performed until the power-depleted load node is marked as accessed, at which point the search terminates and the initial transfer path is output.

6. The method for adjusting the operation mode of a distribution network based on Dijkstra's algorithm as described in claim 5, characterized in that, The cumulative path cost of each adjacent node is constructed based on the aforementioned equivalent weights, specifically including: Based on the power supply branch identifier between the current extended node and each adjacent node, the weight component of the corresponding branch is extracted from the running equivalent weight and determined as the branch path cost of each adjacent node. Read the historical cumulative path cost value recorded by the current extended node, and perform step-by-step superposition of the historical cumulative path cost value and the branch path cost value to generate the cumulative path cost value of each adjacent node.

7. The method for adjusting the operation mode of a distribution network based on Dijkstra's algorithm as described in claim 1, characterized in that, Applying a fixed offset perturbation to the initial transfer path, iteratively calling the Dijkstra algorithm to generate multiple alternative transfer paths, terminating when the currently searched transfer path overlaps with an already generated alternative transfer path, specifically includes: Extract each power supply branch contained in the generated power supply path, and superimpose a fixed offset penalty value on the corresponding operating equivalent weight of each power supply branch to generate a perturbation topology weight map. Based on the perturbation topology weight graph, Dijkstra's algorithm is invoked to perform path search, and the currently searched transfer path is written into the candidate path set; The node sequence of the currently searched transfer path is compared one by one with the node sequences stored in the candidate path set. If there is a completely consistent matching result, the iterative search is terminated. If there is no completely consistent matching result, the currently searched transfer path is written into the candidate path set, and the process of weight perturbation and path search is returned to continue generating new paths.

8. The method for adjusting the operation mode of a distribution network based on Dijkstra's algorithm as described in claim 7, characterized in that, The generation of the perturbation topology weight map involves superimposing a fixed offset penalty value on the operating equivalent weights corresponding to each of the aforementioned electron supply branches. Traverse each power supply branch included in the generated power supply path, and add a fixed offset penalty value to the operating equivalent weight corresponding to each power supply branch one by one to generate the disturbance weight value corresponding to each power supply branch. Each disturbance weight value is written back to the corresponding branch weight field in the topology data structure storing the distribution network topology map, while keeping the equivalent operating weights of the remaining branches in the distribution network topology map unchanged, thus generating a disturbance topology weight map.

9. The method for adjusting the operation mode of a distribution network based on Dijkstra's algorithm as described in claim 1, characterized in that, After adjusting the switch status based on each alternative power transfer path, the distribution network operation status is checked, and the target reconfiguration strategy for the distribution network operation mode is obtained, which specifically includes: The opening and closing status of sectionalizing switches and tie switches in the distribution network topology is updated based on the switching action sequence included in each alternative power transfer path. Power flow calculation is performed to extract the voltage amplitude and branch load rate of each node, and compared with the voltage safety threshold and thermal stability limit threshold to generate a verification identifier. Select qualified candidate transfer paths based on the verification identifier, construct a candidate strategy set, extract the cumulative path cost corresponding to each candidate strategy and sort them in ascending order, and select the candidate transfer path corresponding to the first position of the sequence as the target reconstruction strategy.

10. The method for adjusting the operation mode of a distribution network based on Dijkstra's algorithm as described in claim 9, characterized in that, The process of selecting qualified alternative supply paths based on the verification identifier and constructing a candidate strategy set specifically includes: Read the voltage over-limit flag and branch overload flag in the verification flag corresponding to each alternative power transfer path. When the voltage over-limit flag and branch overload flag of a certain alternative power transfer path are both in a negative state, the verification flag of the alternative power transfer path is determined to be qualified. All candidate supply paths that have been verified as qualified and their corresponding switching action sequences are summarized to form a candidate strategy set.

Images